ai-seo

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly requires checking and analyzing AI answers and competitor pages across public platforms (Step 1: "Check AI Answers for Your Key Queries" — run queries through ChatGPT, Perplexity, Google AI Overviews) and calls out public third-party sources like Wikipedia, Reddit, and review sites, meaning the agent must fetch and interpret untrusted, user-generated web content that can change its actions.